4

On each project we submit drafts for peer review. Whenever one particular colleague of mine performs a review of my work, she always returns it with no comments. I always get comments from everyone else. This leads me to suspect that she is not reviewing it and simply billing time to the project.

I work on engineering problems that provide drinking water to communities - I rely on an effective review process to catch occasional errors and identify potential improvements in my work.

How should I address this suspicion, either directly with my co-worker or with my manager (or someone else)?

In the past I've:

  • Mentioned to my manager that I consistently get no comments from this one individual
  • Mentioned to my co-worker that other reviewers consistently have feedback for me and would appreciate feedback from her
  • 1
    Have you given the same work to someone else to review? Do they come up with issues? – Erik Aug 22 at 14:36
  • 3
    Does your employer have a standard for what a review is supposed to entail, or what a review is supposed to find? If not, it would be no surprise at all that there is some variation due to personal style or personal interpretation of what a "review" is supposed to be. She may be innocently surprised by your implication that she should be finding something. – dwizum Aug 22 at 14:39
  • 2
    What @dwizum said. For some people a review is looking over the changeset, for others raising comments for every infraction of coding style, e.g. blank lines, and for others comparing parts of the commit to the requirements. Sounds like you need to have a meeting to agree on what constitutes a PR. It's all about communication. – Justin Aug 22 at 14:53
  • For these particular projects we are the only two persons that can perform the work. So we always review each other's work. I give her many comments because I want to make sure she puts out quality work. I get no comments. – user107558 Aug 22 at 14:59
  • That's beside the point. Who gets to decide what "quality work" means? How to you measure that? How do you determine if someone else is making quality work? Those questions seem trivial and obvious, but get a dozen developers in a room, who have no pre-agreed upon standard, and ask them simply "review this code" and you'll almost certainly get 12 different answers. You have to know what a review is supposed to mean before you can decide if a given review is "good" or thorough or whatever. – dwizum Aug 22 at 15:16
2

Talk to each other - have a team / departmental meeting to agree on what constitutes a Pull Request (Peer Review, etc).

Maybe you also have different levels of PR, depending on the complexity of the commit.

  • Just review the changes?
  • Compare changes against requirements spec?
  • Has submitter also done minor refactoring to improve the codebase?
  • Are there unit tests?

It doesn't matter as long as you have generally accepted guidelines and ideally more than one approval.

  • 1
    The job expectations are simple. That she perform a review, checking everything. I'm not talking to her about it anymore. She should be replaced. – user107558 Aug 22 at 15:01
  • 3
    @RR2 I'm not talking to her about it anymore. Then I guess you'll never know the answer to your question? – KingDuken Aug 22 at 15:06
  • 1
    @RR2 I'm not talking to her about it anymore. She should be replaced - then what exactly is your motivation for asking this question? It seems like you're not interested in improving the process. Are you trying to get help in somehow "trapping" her, or proving to some third party that she's not doing her job? – dwizum Aug 22 at 15:17
  • 2
    @RR2 you have been given a direct, and very much correct answer: communicate, verbally. Please act professionally and assume good will until proven otherwise. Respect your colleagues. – aaaaaa Aug 22 at 15:18
  • 1
    If someone is not reviewing the work, and just taking the pay, that is not professional. Why shouldn't this be challenged to get a good review. This is an important part of the project process. That is my main question how do I prove otherwise? After a design fails and a forensic analysis traces it back to a mistake that should have been caught in the peer review process? – user107558 Aug 22 at 15:39
1

If you are sure there are issues that should have been caught (sometimes there really are no comments to be made) then it sounds like you have already gone through all of the usual steps by asking her about this and escalating to her manager. I think your managers are the same from the way you have discussed this, but if not then you can escalate to your own manager or whomever is accountable for the work you are doing.

A simple way is to add an obvious error (or multiple errors) into your work and submit it for review. But it doesn't seem a good idea just to try and catch someone out in that way and this may backfire if the manager thinks you are deliberately trying to catch her out

Perhaps by saying to your manager that you found an error yourself after you received the review back with no comments and that the obvious error was not caught, that might work

  • 3
    The other possibility is she's reviewing things "late" and is already seeing comments, assuming it's the kind of review where everyone sees everyone else's comments. As a side note, I was close to down-voting your answer because it also includes the "lay trap for co-worker" suggestion. – Julie in Austin Aug 22 at 16:09
0

Since you're talking about math models for water supply equipment...

There should be a business process for approving important and life-critical technical decisions in your company. In its simplest form, there should be at least a checklist document where each of the experts involved in review should write "checked and approved" and put their signature next to every action they should have performed during the review. That checklist would automatically define the scope of the review.

A better way could be to have another expert double-check the calculations independently, including the choice of parameters and put their signature only if everything matches. But that's another level.

Talk to your boss. It's a business process that they should have established.

  • This is excellent. I wish there was a checklist already. Just a checkbox to say review performed at this point. – user107558 Aug 22 at 16:20
0

The first question here is: Why do you care?

If she missed some errors, this may be a problem, but if right processes at at place, then she will be caught not reviewing your job.

And if you doesn't make mistakes, it doesn't matter if she reviews your work or doesn't, does it?

  • Why, caring for the final outcome for the company is a great personal trait. That's involvement. A good manager strives to get involved employers... – Igor G Aug 23 at 18:43
-2

Lol, how about leaving a trace of comments in your flow to fill out ?

it could be simple as - if you get here write 5x6 result in the next comment line

2-3 times comments are going to be missing - you have empiric proof and can go to your manager with it if comments are there - she does the review

Unless you don`t like the person and looking for a reason to complain about her to your manager :)

Edit: @aaaaaa if , as OP stated, no comments came from this person and no change from verbal communication (request for comments), there should be an indicator to confirm / support his claim that code is not reviewed

@cdkMoose from what i understand OP is responsible for his own code and PR is NOT a formality, there is always a chance of missing something. By not performing a PR that "coworker" is putting OP and his work / status / advancement at risk

@dwizum from the post its apparent that conflict is already happening. No reaction to OP discussions with this coworker, and next stage is to talk to a manager, hollow stipulations are worse than indicators because its always going to be OP word against coworker

@JulieinAustin what is wrong with calling her out if no feedback is given and no reaction to one on ones?

@dwizum i see, the word choice may have been not a best one. i meant it as dictionary does: "a thing that is done simply to comply with requirements of etiquette, regulations, or custom." , Meaning that it is part of methodology in the OPs company

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy